ODK-AI: Coding Ontologies Using AI Tools
ODK-AI is a Docker container for running claude-code (and in future, similar tools) with ontologies. It is designed to be executed either interactively or in "headless" mode.
For more details, see this tutorial (in progress)
The container extends ODK, which means any tool available to ODK (e.g. ROBOT) is available for claude-code
to use.
Ontology Development is a Team Science with many "Agents"
Ontology development is typically a collaborative process involving multiple stakeholders. The diagram below illustrates the traditional workflow:
graph TD
A[Scientists] -->|Write papers| B[Scientific Literature]
B --> C[Curators]
C -->|Curate and extract terms| D[GitHub Issues]
D --> E[Ontology Editors]
E -->|Edit ontology with Protégé| F[Changes to ontology files]
F -->|Push to repository| G[Pull Requests]
G --> H[Ontology Reviewers]
H -->|Review and approve| I[Merged Changes]
I --> J[Updated Ontology]
J -->|Used by| A
style A fill:#f9f,stroke:#333,stroke-width:2px
style C fill:#bbf,stroke:#333,stroke-width:2px
style E fill:#bfb,stroke:#333,stroke-width:2px
style H fill:#fbb,stroke:#333,stroke-width:2px
This team science approach involves multiple human "agents" with specialized roles working together to maintain and expand ontologies.
ODK-AI is designed to help with any piece of this that involves interacting with GitHub
Features
- Built on ODK container with tools like ROBOT already installed
- Automatically initializes template files for your ontology project
- Supports both interactive and headless modes
- Can create and manage pull requests
- GitHub CLI integration for issue tracking
Quick Start
Get up and running with:
See the Getting Started guide for more details.
Caveats
- Workflow is primarily tested with ontologies that keep their source in
.obo
format - This project is in early development and may have bugs